Data management has gained interest in spine care and large-scale datasets have improved the approach to clinical practice and research. Multicenter registers have enabled more discriminant analysis and prediction of clinical patient-related outcome measures and risk assessment in spine surgery. Analyses are traditionally based on statistical methods. Modern concepts of prediction include machine learning, which represents a subgroup of artificial intelligence. They originate from computer science and use algorithms established from previous data analysis to accomplish certain tasks. Different machine learning methods, such as decision trees or support vector machines, are used for clinical prediction models. They can be used as decision-making tool to help clinicians in complex fields where machine learning algorithms have previously learned from large clinical databases. Artificial neural networks and convolutive neural networks process combinations of information used in image or voice recognition. The main scientific and clinical applications that have been tested are diagnostic spinal imaging, segmentation of intraoperative images for surgical navigation, including augmented reality, or robotics. Deep learning combines a set of machine learning methods, which is used for modeling complex relationships with a high degree of abstraction. This technique has a limited use in spine surgery today but might be implemented in the future.